Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d270Policy30d65, 13 pendingSources1,015

Gemini 3 Pro

GoogleReleased Nov 18, 2025Current

Score basis

Gemini 3 Pro is at the frontier in multimodal tasks; strong in long context; capable in reasoning, instruction following, and factuality; and behind the leaders in coding and agentic tasks. Too few results yet to rate safety, math, or multilingual tasks.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Frontier1 capability
  1. MultimodalLeads5 of 6
Strong1 capability
  1. Long Context−8.6%3 of 3
Capable3 capabilities
  1. Reasoning−18.0%5 of 6
  2. Instruction Following−21.6%2 of 3
  3. Factuality−23.2%4 of 4
Limited2 capabilities
  1. Coding−26.4%7 of 10
  2. Agentic−38.6%3 of 7
Not rated3 capabilities

Too few results yet to rate Safety, Math or Multilingual.

Price

$2.00input$12.00outputper million tokens

From Artificial Analysis · All prices

Costs more than 85% of 330 priced models · 3:1 input-to-output blend, log scale

Evidence

194results on129benchmarks

  • 41 independently verified
  • 31 aggregator
  • 28 vendor-reported
  • 94 cross-referenced

From 40 sources · latest Oct 8, 2026 · How verification works

Research

36 papers reference Gemini 3 Pro

Gemini 3 Pro benchmark results

194 results on 129 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Multimodal

FrontierFull multimodal ranking

Leads the field5 of 6 ranked benchmarks measured

Show 6 more multimodal resultsHide 6 multimodal results

Long Context

StrongFull long context ranking

8.6% behind the leader3 of 3 ranked benchmarks measured

Show 2 more long context resultsHide 2 long context results

Reasoning

CapableFull reasoning ranking

18.0% behind the leader5 of 6 ranked benchmarks measured

Show 10 more reasoning resultsHide 10 reasoning results

Instruction Following

CapableFull instruction following ranking

21.6% behind the leader2 of 3 ranked benchmarks measured

Show 2 more instruction following resultsHide 2 instruction following results

Factuality

CapableFull factuality ranking

23.2% behind the leader4 of 4 ranked benchmarks measured

Show 2 more factuality resultsHide 2 factuality results

Coding

LimitedFull coding ranking

26.4% behind the leader7 of 10 ranked benchmarks measured

Show 15 more coding resultsHide 15 coding results

Agentic

LimitedFull agentic ranking

38.6% behind the leader3 of 7 ranked benchmarks measured

Show 5 more agentic resultsHide 5 agentic results

Math

Not enough dataFull math ranking

0 of 5 ranked benchmarks measured

Show 3 more math resultsHide 3 math results

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 111 more resultsHide 111 results

Gemini 3 Pro: common questions

Who makes Gemini 3 Pro?

Gemini 3 Pro is made by Google.

When was Gemini 3 Pro released?

Gemini 3 Pro was released on Nov 18, 2025, according to Artificial Analysis.

What is Gemini 3 Pro good at?

Gemini 3 Pro is at the frontier in multimodal tasks; strong in long context; capable in reasoning, instruction following, and factuality; and behind the leaders in coding and agentic tasks. Too few results yet to rate safety, math, or multilingual tasks.

How much does Gemini 3 Pro cost?

Gemini 3 Pro costs $2.00 per million input tokens and $12.00 per million output tokens, according to Artificial Analysis. At a mix of three input tokens to one output token, it costs more than 85% of the 330 priced models we track.

How many benchmarks has Gemini 3 Pro been tested on?

We track 194 results for Gemini 3 Pro on 129 benchmarks from 40 sources, 41 of them independently verified. The latest was recorded on Oct 8, 2026.

About this record

Where Gemini 3 Pro's numbers come from, and every name it appears under.

Tracked since
Apr 25, 2026
Newest source mention
Aug 29, 2026

Where the results come from

Verification: 194 scores · 41 independently verified · 31 aggregator-attributed · 94 vendor cross-reference · 28 vendor-reported. How these tiers are assigned

From 40 sources on 16 sites. Hugging Face supplies 87 of them; the 41 independently verified results come from 11 sites. Bars are coloured by trust tier.

  • huggingface.co87
  • artificialanalysis.ai31
  • api.llm-stats.com16
  • deepmind.google12
  • matharena.ai9
  • raw.githubusercontent.com7
  • www-cdn.anthropic.com7
  • storage.googleapis.com6
  • swebench.com5
  • labs.scale.com3
  • 99franklin.github.io2
  • arcprize.org2
  • datasets-server.huggingface.co2
  • epoch.ai2
  • lmarena.ai2
  • simple-bench.com1

Also known as

How our sources name Gemini 3 Pro at each reasoning setting.

SettingShort formAPI id
lowgemini 3 pro preview (low)gemini-3-pro-low
highGemini 3 Pro Preview (high) Gemini 3 Pro Thinking (High)—
Also listed asGemini 3.0 Progemini-3-pro-previewgemini 3 pro preview (2025-11-18)gemini 3 pro (preview)gemini 3 pro (refine.)gemini 3.0